Numerous human tumor lines do not express the HLA class II antigen presenting molecules in response to IFN-gamma (interferon-gamma), a defect that may be related to tumor cell escape of anti-tumor immunity. In some cases, this defect has been attributed to loss of the retinoblastoma protein. In many other cases, lack of HLA class II induction is attributable to lack of induction of class II transactivator (CIITA), which is required for HLA class II activation by IFN-gamma. In some cases, CIITA noninducible pancreatic tumor lines are defective in the function of interferon regulatory factor-2 (IRF-2). Preliminary studies for this proposal strongly indicate that IRF-2 is important for the production of CIITA, and in particular that IRF-2 function involves cooperation with yet another factor, IRF-1, previously implicated in the production of CIITA. This proposal is designed to understand the molecular basis for why IRF-1 and IRF-2 cooperate to effect the IFN-gamma response of the CIITA promoter. The proposal will also investigate the basis for lack of IRF-2 DNA binding function in human pancreatic tumor lines, as well as other IRF-2/CIITA related anomalies seen in several other tumors lines. Results from this work will provide a better understanding of how multiple factors interact to effect gene activation. In particular, this work will help resolve the bases for functional differences between the closely related transactivators, IRF-1 and IRF-2, and their response elements. In the case of the response elements, this work will likely establish that minor differences generally considered to be irrelevant variations of a "uni-functional" consensus sequence actually have important functional consequences. This proposal also will likely lead to a better understanding of gene regulation relevant to autoimmunity and tumorigenesis. It is also likely that this work will identify mutations in IRF-2 that could serve as pancreatic tumor markers. If so, this would be the first demonstration of a role for IRF-2 mutations in human tumor. Methods: EMSA, in vivo footprinting, DNA transfection, DNA sequencing, DNA cloning.